Maryam Zaheer

Maryam ZaheerFull-Stack Developer, AI/ML & Computer Vision

Recent project: MedInsight: Evaluating Retrieval-Augmented Vision-Language Models for Evidence-Grounded Medical Image Understanding

Building systems
that solve real problems.

I'm Maryam, a Computer Science student passionate about building AI-powered and full-stack software that solves real-world problems—from intelligent web applications to trustworthy machine learning systems.

Lahore, Pakistan · maryamzaheer2006@gmail.com
Research
2026

MedInsight: Evaluating Retrieval-Augmented Vision-Language Models for Evidence-Grounded Medical Image Understanding

Does giving a medical vision-language model retrieved similar-case evidence actually make it more accurate — or does it just look like it should?

Python
PyTorch
BLIP-2
CLIP
FAISS
Retrieval-Augmented Generation
VQA-RAD test exampleimage + clinical questionBaseline pathno retrieval usedQuestion-only promptno evidence blockRetrieval-augmentedCLIP image encoderFAISS index searchROCOv2 · ~60K pairs · cosine simPrompt + evidencetop-k captions prependedShared generation — BLIP-2greedy decode · identical for both conditionsAnswer + scoring · exact match · BLEU-4 · ROUGE-L